Background Through the Supportive Services for Veteran Families (SSVF) program, Veterans and their families who are experiencing homelessness or at risk of becoming homeless can receive of financial support and other services from private non-profit organizations with the goal of stabilizing their housing. The impact on Veteran housing, healthcare cost, and health outcomes as well as Veteran satisfaction with this program is not known. Specific aims Aim 1: Describe geographic variation in the amount of SSVF services provided and identify predictors of this variation for SSVF recipients for FY2012-FY2017 using data from the HMIS. Aim 2: Perform an outcome and economic assessment of the SSVF program. 2a: Measure the impact of the rapid re-housing component of SSVF on housing, health, and healthcare cost outcomes through retrospective cohort analyses for FY2012-FY2017. 2b: Conduct grantee-specific and overall economic evaluations of the SSVF program compared to usual care using input parameters from the published literature and from Aims 1 and 2a. Aim 3: Identify barriers and facilitators to SSVF use and Veterans' satisfaction with the program using semi- structured interviews with VA operations partners, SSVF grantee employees, VA homeless services providers, and Veterans using SSVF services. Significance of the proposed research and relevance to Veterans' health The loss of stable housing has a devastating effect on Veterans' health and well-being. In addition, homelessness can be costly both to the healthcare system and society as a whole. Solving the problem of Veteran homelessness is one of the most important initiatives the VA is currently pursuing. Unique features and innovation of the project Our study ? comprising over 400 sites within the US ? will be by far the largest study of a homeless intervention ever conducted in the US. In addition, we will use both structured and unstructured data and detailed, Veteran-level data on services and financial assistance received by Veterans enrolled in this program. Methodology to be used In Aim 1 we will construct multi-level regression models to identify both individual- and community-level predictors of SSVF services and expenditures. In Aim 2a, we will fit outcome models relating SSVF program initiation to the subsequent mortality, exit from homelessness, and cost outcomes in weighted populations defined by propensity scores. In Aim 2b, we will construct simulation models using input parameters from a variety of sources. And in Aim 3, we will conduct interviews with national VA stakeholders, SSVF grantee employees, VA homeless service providers, and Veterans who have enrolled in the SSVF program. Expected results In Aim 1, we expect to find considerable variation in services and expenditures across the country, which will be explained by observable Veteran- and community-level characteristics. We anticipate that our Aim 2a analyses will find that enrollment in SSVF will lead to increased rates of stable housing, lower mortality rates, and greater outpatient healthcare costs but lower emergency department and inpatient costs. in Aim 2b, our hypothesis is that the SSVF program will be cost-effective overall and yield a positive return on investment compared to usual care but these effects will vary substantially across SSVF sites. Finally, our qualitative data from Aim 3 will provide insight into ways to improve the SSVF program, enhance implementation strategies, and explore the experience of Veterans with housing instability.